Features · AGI Future Foundation
The Windfall Promise
There is a sentence in the Foundation’s pitch deck that is easy to applaud and hard to keep. It says that everything above a certain threshold belongs to humanity. The sentence is not a donation pledge. It is a legal pre-commitment, hardwired into governing documents before the profits exist, so that sharing is the default rather than a decision a future board must find the courage to make. The trouble is that the sentence, as it stands, has no number in it. No threshold percentage. No measure of global GDP. No definition of profit. No share. No administrator. No auditor. This feature is the full deep dive into that promise: what it commits to, what it leaves blank, why the blanks matter, and why a promise with no number in it is still worth making, as long as nobody pretends the number is already there.
The promise made before the money
The Windfall Clause begins with a diagnosis, and the diagnosis is about human nature under pressure. Voluntary corporate philanthropy, the pitch deck argues, often fails under immense market pressure. A board that means to share profits later can be sued, outvoted, acquired, or worn down before the profits arrive, and by the time the money is real, the people who promised to share it are often no longer the people deciding what happens to it. The clause answers that by making the sharing decision before the profits exist, hardwired into the governing documents, so that sharing is the default. The pitch calls it a legal pre-commitment, and the legal paper built to the pitch standards keeps that word, because the timing is the whole of the idea. A Windfall Clause adopted after a series is profitable, or amendable by a simple board majority at the moment profits arrive, does not meet the pitch standard. The promise has to be made when it costs nothing, so that it holds when it costs everything.
The obligation, as the pitch states it, runs to the 33 W3 LLCs, the profit-generating series, not only to the Foundation or to a benefit corporation at the top. That is a deliberate choice, and the legal paper flags it as a key point: a Windfall Clause that lives only in a parent company’s articles, with no series-level obligation, does not meet the pitch standard. The clause must appear in each series’ designation documents, not only in the parent’s articles, and it applies to each of the 33 W3 series listed in the April 2026 internal list, and to any series added after adoption, unless counsel expressly excludes a series in its designation document, with the default being inclusion. Today it has no series to bind. No W3 series has been designated, no series designation document exists, and the Delaware cohort master that would hold the new series has not been verified as filed. The clause is a draft waiting for something to attach to.
The clause sits inside a larger mechanism of three interlocking gears, and the pitch names them on its fourteenth slide: the PBC Shield, the legal protection for the safety mission; the Windfall Clause, the distributive engine for global wealth; and the OGI Shares, the economic inclusion that defuses conflict. The Shield is the legal answer to the fear that a board will be sued for sacrificing near-term profit for safety: California public benefit corporation status, which rewrites the board’s duties so that safety and public benefit are part of the job, with directors required to consider the benefit purpose alongside shareholder interests under sections 14620 and 14621, and an annual benefit report under section 14630. The Shield is expanded business-judgment protection for benefit-considering decisions, subject to those duties. It is not immunity. The pitch deck’s sixth slide says the board becomes legally immune to shareholder lawsuits when sacrificing short-term commercial profits for safety guardrails, and that overstates California law: a benefit corporation’s directors are not immune from shareholder suits, and a PBC can still be sued, acquired, or have its benefit purpose amended by the required vote. The legal documents correct the claim, and this feature repeats the correction, because a promise built on an overstatement will not survive its first test.
The close, and the strategy behind it
The deck closes, on its fifteenth slide, with a sentence that sounds like a benediction and reads like a contract: prosperity, distributed, and the wealth of a new world not to be hoarded but to be shared, through the Open Global Investment model and binding benefit sharing, so that the advent of AGI is not the final zero-sum battle of human history but the foundation of its most extraordinary era. The eleventh slide makes the same point as strategy rather than sentiment: Open Global Investment means a low incentive to race, because participation in the economic upside is guaranteed, a vested interest in the Foundation’s secure and safe success, and a cooperative, stabilised transition to post-scarcity, against a status quo of high incentive to race driven by fear of exclusion and dependency, espionage and sabotage and rushed deployment, and a high risk of existential conflict. The twelfth slide extends the logic to rivals: by allowing foreign nations and international rivals to participate in the profits of the 33 W3 companies, the calculus changes, so that destroying or recklessly racing the Foundation transforms from an act of national security into an act of profound economic self-destruction. That is the promise at its most ambitious, and it is also the promise at its most fragile, because it depends on the participation being real, the screening being honest, and the clause being in force before the profits arrive. A promise that depends on all three is a promise that needs all three written down.
The threshold that is not a number
The threshold is the heart of the clause, and it is also the part of the clause that is least like a threshold. The pitch expresses it as a percentage of Global GDP, not as a fixed dollar amount, and the tenth slide draws the line plainly: profits up to the threshold reward early investors and sustain research and development, and everything above the threshold belongs to humanity. Below the threshold, profits follow the normal rules, with standard returns to investors and retained funding for research and development. Above it, the excess becomes Windfall Profits, paid to a Windfall Administrator for distribution to the four Benefit Buckets. The shape of the idea is clear. The number is not there.
Four blanks stand between the shape and a calculable trigger, and the legal paper lists them as brackets, not terms. The percentage itself is blank. The GDP measure is blank: which measure, World Bank or International Monetary Fund, which year, which vintage. The profit measure is blank: audited net income, operating profit, or distributable profit, after tax, after a research and development reserve, with counsel to choose one and define it. And the scope is blank: each series separately, the W3 aggregate, or the PBC plus W3 aggregate. Until those are defined, a percentage of Global GDP with no percentage, no measure, and no scope is a direction of travel, not a trigger, and the clause must not be described to an investor as triggered at a dollar figure. The measurement period is blank too: annual or cumulative, with counsel to choose. The auditor who calculates the threshold is named only by role, not by person, until appointed.
It is worth pausing on why the pitch chose a percentage of Global GDP at all, rather than a round dollar number, because the choice is doing real work even while the number is missing. A fixed dollar threshold would be overtaken by inflation, by growth, and by the sheer scale of the economy the clause is meant to measure itself against. A percentage of global output scales with the world it is meant to serve, and it says something about the ambition: the trigger is not a company getting rich, but a company becoming large relative to the whole world economy. That is also why the blank is dangerous. A percentage sounds precise, and a reader who hears a percentage of Global GDP will assume a number exists somewhere, in a footnote or an appendix, waiting to be found. There is no number. The paper is candid about that, and the candour is the point. A threshold with no percentage is not a lie. A threshold presented as if it had one would be.
The scope blank is the one that will cause the most argument when the clause is finally drafted, because it decides whose profits count. Three answers are on the table, and the draft lists them without choosing: each series separately, the W3 aggregate, or the PBC plus W3 aggregate. Each series separately means a single wildly profitable series triggers the clause on its own, even if the family as a whole is still losing money, which is the strictest reading and the one most faithful to the pitch’s claim that the obligation runs to the 33 series. The W3 aggregate means the family’s profits are pooled for the measurement, so that a profitable series can be offset by a loss-making one, which is gentler on the family and harder on the promise. The PBC plus W3 aggregate folds the benefit corporation’s own profits into the same measurement, which is the broadest reading and the one that most closely matches the pitch’s talk of the wealth of a new world. Counsel must choose one, and the choice must be stated in the clause, because a trigger that can be read three ways is not a trigger at all. It is a negotiation waiting for the day the money arrives, which is exactly the day a negotiation will be hardest to win.
The four buckets, and what they are for
Above the threshold, the excess goes to four Benefit Buckets, and the buckets are the most concrete part of the whole clause, because they name where the money goes rather than how much of it there is. The buckets are Global Public Goods, Healthcare Access, Energy Infrastructure, and Technological Self-Determination, in equal shares unless counsel or the board sets different shares, in each case for projects advancing the UN Sustainable Development Goals, directly funded rather than only through intermediaries, as the Administrator’s published criteria provide. The pitch’s tenth slide shows three of them — Global Public Goods, Healthcare, and Infrastructure — and its thirteenth slide names Healthcare Access, Energy Infrastructure, and Technological Self-Determination, with the ultimate goal stated as the global actualisation of the UN Sustainable Development Goals, bypassing traditional bureaucratic bottlenecks to directly fund post-scarcity infrastructure. The four-bucket list in the legal paper combines both slides, and this feature follows the legal paper.
The buckets are worth reading one by one, because each one answers a different fear about what concentrated AGI wealth would do. Global Public Goods is the bucket for the things no single country or company will pay for on its own: the shared infrastructure of knowledge, safety, and coordination that a world with powerful AI will need. Healthcare Access is the bucket for the most immediate human need, and the one where the gap between what exists and what is possible is most visible. Energy Infrastructure is the bucket for the physical foundation everything else runs on, and the one where the pitch’s talk of post-scarcity infrastructure is most literal: power is the precondition for everything else the clause promises. Technological Self-Determination is the bucket for the right of peoples and countries to shape their own technological future, rather than depending on systems built elsewhere, under rules they did not write. Together they describe a world in which the benefits of AGI are distributed, not hoarded, and in which the distribution is direct, not mediated by the institutions the pitch says it wants to bypass.
Two guardrails attach to the buckets, and both are written into the draft clause as prohibitions rather than aspirations. No Windfall distribution may be made to a shareholder, director, officer, or series manager as a return on their shares or interests, and no distribution may be used to buy control, influence a government, or fund a political campaign. Counsel must confirm charitable and public-benefit eligibility per bucket and per recipient, because a bucket is not a licence to spend on anything that sounds good. The Windfall Share itself — the portion of excess profits that goes to the Administrator — is blank, and the pitch calls it a massive portion without giving a number, and none may be quoted as the share. The remaining portion of excess profits, if any, follows the normal distribution rules, and counsel must confirm that the Windfall Share plus the remainder equals one hundred percent of excess profits, with no gap. The investor disclosure the clause requires in every offering document that mentions it states the bargain plainly: the Windfall Clause is a pre-commitment to share a blank percentage of profits above a blank percentage of Global GDP with humanity, through the Benefit Buckets, and it reduces the profits available to shareholders above the threshold, by design. It is not a guarantee of any return, and it does not guarantee that any profit will ever exceed the threshold.
The administrator, the auditor, and the empty chairs
A promise needs a person to keep it, and the Windfall Clause names three roles that are still empty. The first is the Administrator, who receives the Windfall Profits and distributes them to the buckets. The draft leaves the choice open: the Foundation as a nonprofit, if separate, a Windfall Fund, or a trust, with counsel to advise on tax and charitable status. If the Administrator is the nonprofit Foundation, Windfall money is held in a segregated Windfall account, reported separately from Foundation donations, and used only for the Benefit Buckets. It does not become general Foundation funds, and it does not fund AGI Corp, a series, or a SAFE round. That separation matters, because Windfall money arriving in a nonprofit’s general account would be indistinguishable from a donation, and a donation can be spent on the donor’s purposes, while Windfall money can only be spent on the buckets. The Administrator’s reporting is an annual public Windfall report, covering the threshold calculation, the excess profits, the Windfall Share paid, the bucket allocations, the recipients, and the audit, with the calendar in the compliance paper.
The second empty chair is the auditor, who calculates the threshold and the excess profits from audited accounts, and publishes the result with the GDP measure, the vintage, and the profit measure stated, so that a reader can recompute the trigger. The draft names the role, an independent auditor or certified public accountant, and not a person, until one is appointed. The third is the dispute forum: disputes over calculation go to arbitration or to court, as counsel sets, and they do not suspend the obligation to pay the undisputed portion. The payment timing is blank too: the Windfall Share is to be paid within a stated number of days of audited accounts for the period, with the number of days blank for counsel to set. None of these blanks is a detail. A clause with no administrator has nobody to pay. A clause with no auditor has nobody to calculate. A clause with no payment timing has no deadline. The draft is honest about all three, and the honesty is what makes it a draft rather than a decoration.
The SmartCiti.X corporation plan shows how the clause would attach to a real company, if one is ever formed. The Windfall Clause would be incorporated into the corporation’s charter and bylaws and into every SAFE and preferred agreement for the corporation, with the threshold as a percentage of global GDP, the Windfall Share, the administrator, and the segregated account all blank for counsel, and survival on a sale, merger, or change of control to be confirmed in writing by counsel rather than assumed. The scope there is the corporation’s own profits, unless counsel expressly expands it to affiliates. That is the corporate version of the series-level adoption the pitch requires, and it is the clearest illustration in the papers of the clause’s central discipline: the promise must live where the money is made, in the entity that earns the profit, not only in the parent that holds the control.
The lock: supermajority, Class C, and the open door
The amendment protection is the part of the clause the pitch calls the unbreakable mechanism, and it is the part that has to work, because every other part of the clause can be undone by a vote at the moment the profits arrive. The draft provides that the clause may be amended, suspended, or repealed only by a supermajority vote, with the threshold to be set by counsel, for example a stated percentage of Class C plus a stated percentage of Class A and B voting as separate classes, plus Foundation board approval, chosen to make amendment possible in a genuine emergency but not by a simple board majority at the moment profits arrive. A purported amendment that reduces the Windfall Share, raises the threshold, or removes a Benefit Bucket, adopted without that vote, is void. Class C’s 1,000-vote control is part of this protection, and Class C holders must disclose that control to Class A and B investors as a governance fact, not as a guarantee that the clause will never change.
Class C is the key to the lock, and it needs explaining, because it is also the most concentrated power in the whole structure. The share structure built to the pitch standards has three classes. Class A carries one vote per share and full economic participation: the global public’s class, open worldwide subject to securities law, the class that carries almost entirely profit participation and a small amount of voting power. Class B carries ten votes per share, is both economic and voting, and is reserved for aligned partners and sovereign entities meeting strict safety and geopolitical frameworks, screened for sanctions, export controls, and national-security review. Class C carries 1,000 votes per share, is held exclusively by the Foundation, and has no economic participation at all: control only. The pitch calls that absolute, unshakeable ethical control, and the legal paper is careful to translate it: Class C, even in small numbers, dominates Class A and Class B on any ordinary vote, and that control must be disclosed to Class A and B investors as a governance risk, because they own economics and do not control safety decisions. Control is not ownership. The Foundation’s control over the 33 series is exercised through Class C control at the benefit-corporation level, through mission and voting agreements between the Foundation and each series, and through each series’ Steward Council mission gate, not by the Foundation owning each series’ economics.
Open Global Investment is the open door that the lock protects, and the two belong together. The pitch’s second and third slides frame it as a transformation: AGI must not be a zero-sum geopolitical weapon, and Open Global Investment turns it from a closely guarded national asset into a globally accessible economic engine. Broad ownership, widely open to international investment, drastically reducing expropriation risk. Ethical guardrails, operating within a framework of rigorous safety, guided by a specialised corporate hierarchy. Strategic inclusion, inviting rival nations to participate in the upside, removing the urgency of independent, unsafe AI programs. The strategic logic is deliberately counterintuitive: a closely guarded national asset gives every capable party the same incentive, to race to build it first, keep others out, and accept safety risks rather than fall behind, while inviting rivals into the economic upside means undermining the enterprise would mean undermining their own stake. It is unproven, with no terms, no participants, and no track record today. The participation instrument is blank. Eligibility is blank. Price and valuation are blank. Governance rights are blank, with a default of none. Transfer and liquidity are blank, with no liquidity, secondary market, or buyback promised. Class A holders are told upfront what they are buying into: economics, not control.
Two cautions belong in the same passage, because the pitch’s rhetoric runs ahead of its evidence in both. The first is the Gini figures: the eighth slide claims that Open Global Investment achieves a significantly fairer global distribution, with a Gini of around 0.89, against a United States-only nationalised AGI programme yielding a highly unequal distribution, with a Gini of around 0.96. Those figures are pitch illustrations, not sourced statistics, and they must not be quoted to an investor, donor, or regulator as measured data. If they are used at all, they must be labelled as a pitch illustration, not a measured forecast. The second is sovereign participation. Class B is restricted to aligned partners and sovereign entities meeting strict safety and geopolitical frameworks, and a sovereign that does not meet the framework does not get Class B, and does not get Class B voting power by buying Class A. Any sovereign or rival participation needs sanctions screening, export-control and foreign-investment review where applicable, and tax and securities advice per jurisdiction. The pitch’s talk of co-opting rivals must never be presented as rivals can buy in without that screening. Class A is the broad-access class. Class B is the restricted class, and the distinction must be kept in every document.
What exists today, what is planned, and what is still missing
What exists today is a pitch deck, a draft clause, and a governance page. The deck is Architecting Global AGI Prosperity, fifteen slides, Part 3, read on 2026-11-03, and the review that verified it notes that the file is byte-identical to an earlier upload under a different name, the same deck with a new filename. The draft clause is sections W.1 through W.7 of the Windfall Clause and Open Global Investment paper, built to the pitch standards, with every operative term in brackets. The governance page on this site carries the clause in full, with the same blanks. What does not exist is everything that would make the clause binding. No series has adopted it. No corporation has adopted it. No threshold, share, administrator, auditor, or dispute forum has been set. No Windfall account has been opened. No Windfall report has been published. No profit has ever exceeded any threshold, because no threshold has ever been set, and no series has ever been profitable.
What is still missing is a checklist, and the legal paper writes it as one, so that nothing on it can be quietly dropped. The threshold percentage, the GDP measure, the profit measure, and the measurement scope, per-series versus aggregate. The Windfall Share percentage. The administrator entity, the segregated account, and the reporting criteria. The Class C holder, transfer, and conversion rules. The Class B framework, with its safety and geopolitical criteria, its screening, and its qualified-holder definition. Series-level adoption, in each of the 33 series’ designation documents, not only in a parent company’s articles. Counsel’s confirmation that a California benefit corporation and Wyoming series LLCs can each lawfully adopt a clause of this form, and that it survives a sale, merger, or change of control, or, if it does not survive automatically, what successor obligation is required. And securities disclosure, covering the Class A, B, and C rights, the Windfall cost, the Class C control, the sovereign-participation screening, and the labelling of the Gini figures as pitch illustrations, in the offering documents, not only in the pitch deck.
The open question underneath the checklist is the one the clause was written to answer, and it is still open. Can a promise made before the money exists survive the arrival of the money? The clause’s answer is a structure, not a sentiment: adopt it before profits, lock it behind a supermajority and Class C, put it in every series’ documents, and report on it in public, every year, with numbers a reader can recompute. The structure is unbuilt, the numbers are blank, and the series do not exist. What exists is the promise, written down, with its blanks visible, and a record of what would have to be true for it to hold. That is less than a guarantee, and it is more than a slogan. It is a promise that knows what it still needs, and says so.
What this is / What it isn’t
What this is
- A legal pre-commitment, hardwired before profits: the Windfall Clause obligates the 33 W3 series to share a portion of excess AGI profits toward international benefit sharing once a threshold set as a percentage of Global GDP is met.
- Four Benefit Buckets for the UN Sustainable Development Goals: Global Public Goods, Healthcare Access, Energy Infrastructure, and Technological Self-Determination, directly funded, with no distribution to shareholders, directors, or officers as a return.
- A lock: amendment only by supermajority plus the Class C vote, with Class C held exclusively by the Foundation, 1,000 votes per share, control only, no economic participation.
What it isn’t
- Not in force: no series and no corporation has adopted the clause, and it has no series to bind today, because no series has been designated.
- Not a number: the threshold percentage, the GDP measure, the profit measure, the scope, and the Windfall Share are ALL blank for counsel, and none may be quoted as a term.
- Not immunity and not a guarantee: the Shield is expanded business-judgment protection, not immunity; the Gini figures are pitch illustrations, not measured data; and participation in upside, if there is upside, is not a guaranteed return.
Where to see it
The full clause is on the Windfall Clause governance page, with the Shield on the Fiduciary Shield page, the three classes on the Share Structure page, and the open door on the Open Global Investment page. The earlier feature on the same mechanism is The Shield and the Windfall, feature 07, which covered the Windfall Clause as one of three gears; this feature is the full deep dive. The Shield’s repository is public on GitHub: FiduciaryCorporateShield, described there as an agent based fiduciary corporate shield, with a README of two lines, title only, and no substantive code reviewed behind the Shield, which lives in the legal and governance documents, not in code.
Nothing on this page is legal, tax, or financial advice. The Windfall Clause is a draft design with every operative term blank for counsel to set, and it is not in force. The pitch figures cited here are pitch illustrations, not verified data.